Dr. Stefanie Feih can be described as a citizen of the world, based on all the places she has studied, worked and lived. She grew up in Germany with a love of STEM topics, got her M.E. in her native country but also participated in an international exchange program with Cornell University in the U.S. She pursued her Ph.D. in Engineering at the University of Cambridge, U.K., in collaboration with TWI (The Welding Institute), and began a research career that took her from Denmark to Australia with a strong focus on design of lightweight materials (both composite and metal) for wind, aerospace and maritime applications. Dr. Feih joined the Singapore Institute of Manufacturing Technology (SIMTech) in 2014 to work in a more industry-focused research environment. Singapore has strong ties to the Aerospace OEM and MRO industry, a sector that drives exploration and implementation of weight-saving material strategies. Additive manufacturing is a key topic of study for Dr. Feih and her colleagues.

What kinds of projects is your team at SIMTech involved with?

SIMTech develops high-value manufacturing technology and human capital to enhance the competitiveness of Singapore’s manufacturing industry. We collaborate with multinational and local companies in the precision engineering, electronics, semiconductor, medical technology, aerospace, automotive, marine, logistics and other sectors. SIMTech is a research institute of A*STAR.

The additive manufacturing (AM) team at SIMTech works on a wide range of AM technologies for both polymer and metallic components. The team focuses on (i) fundamental understanding of 3D AM technologies, (ii) exotic and nonproprietary material development for use in 3D AM systems and (iii) new applications for 3D AM technologies to complement or replace conventional manufacturing processes. SIMTech also supports local SMEs and multinational industry clients with AM technology adaption solutions and provides industry training courses for AM technology.

How long have you been an Abaqus user?

I have used Abaqus software for the past 20 years for materials research and have also taught FE theory in both university and industry courses. In Singapore, SIMTech staff have participated and presented at the local SIMULIA user conferences over the last eight years to highlight our research outcomes and industry solutions. Abaqus has the ability to conduct highly nonlinear analysis, both in explicit and implicit formulations, and offers a wide range of material and failure models along with the option to program customized user subroutines to interact with the main code. This makes the software a very versatile research tool used widely by both research and industry organizations.

How does your study of lightweight composite materials inform your current work on additive manufacturing technology?

We are currently working on lightweight materials that combine a metal AM lattice core with a fiber-reinforced composite skin. The numerical work on the AM structural investigation is done with Abaqus, and we have published extensively and also presented at the Abaqus user conference in Singapore on this topic (see Figure 1).

Metal AM has the potential to create lightweight structures that exceed the performance of traditionally manufactured composites, but this approach is currently limited by printing size and printing accuracy—hence our paper addressing these parameters, delivered to Science in the Age of Experience this past May (see Figure 2).

SIMULIA’s new AM process simulation framework allows for separate modeling of solid material, powder bed and platform as well as the evolving heat-transfer surfaces of a part as it is built. Why is it important to take all of these components of a build into account when creating a simulation? How can simulation help AM users with quality assurance and cost control?

It is necessary to develop a systematic understanding of factors potentially influencing the print results. Due to time constraints during the simulation, most researchers currently neglect the influence of the surrounding powder bed and platform. It is also unclear whether neighboring parts influence the cooling rates within the metallic structure and what distance should be kept between individual parts. Finally, it is also difficult to establish realistic thermal properties for the powder bed for a large range of temperatures. In our simulation research we showed that the surrounding powder bed material influences the thermal cooling rates of the build part, and the platform, which is normally pre-heated prior to build, contributes as a finite heat sink to the temperature distribution within the build part. Both are therefore important to consider, and the platform should always be modeled. The powder influence may be accounted for by adjusting the film coefficient governing surface convection of the print part to simplify the model complexity.

For how long, and in what ways, have SIMTech and Dassault Systèmes SIMULIA been cooperating in developing simulation capabilities for additive?

Dassault Systèmes and SIMTech started collaborating on additive manufacturing simulation three years ago following a seminar from Dassault Systèmes at SIMTech on the new AM capabilities. We quickly determined a mutual interest in terms of validation and prediction of distortion of our printed parts. Especially with larger build platforms, print failure is one of the greatest cost factors in terms of time and material waste, hence enhancing our simulation capabilities to allow optimization of print orientation, support structures and print parameters for parts with minimum distortion and print defects is a key priority.

Your paper, delivered at Science in the Age of Experience 2017, compared different ways to account for the moving heat source (see figure 3) and also looked at the sensitivity of solution to the mesh. You also did a study on using HPC to scale your simulation times. Are you developing best practices so that you can deliver recipes for AM simulations and make it easier for others in your organization to leverage?

Simulation time is currently the biggest obstacle for AM process simulation. The new concentrated heat-flux model approach developed by SIMULIA allows for multiple print layers to be simulated within one element, which reduces computational time considerably. Elements are then partially activated every time the laser path passes through an element. In contrast to this we also have the more established Goldak-type heat source models, which require very fine mesh resolution for each print layer and hence much longer simulation times. Abaqus offers both approaches within the same modeling framework. From a research point of view it is therefore important to establish validated outcomes for the concentrated heat flux model in terms of predicted temperature and stress distributions, hence allowing for significant reduction in simulation time.
Long-term we would then like to train our engineering staff operating AM equipment to use these available tools prior to print scheduling for print set-up validation.

What’s on your “wish list” for additional simulation functionality? How do you think this would connect to the overall design and manufacturing and final build workflow?

Ideally I wish to link optimized part design (i.e., topology optimization) to process simulation. During the design process, we need to determine best geometry with best print orientation, including optimized support structures. We then envisage potential compensation of the initial geometry via process simulation to allow for near net-shape printing. Process simulation also needs to interface with AM printing equipment to allow unique machine parameters, such as heat source movement, speed and power, to be used as input for the simulation. While work has started on all these individual stepping stones, a unified concept with a reasonable simulation time frame for industry projects has so far not been realized. The one-stop solution remains an ambitious task for software developers and the AM community.

You previously worked on light-weighting composites. How does AM fit into the light-weighting and/or parts consolidation narrative and why is this important for the future of so many industries?

AM offers the opportunity to design and manufacture highly complex and integrated parts, hence reducing the need for costly assembly processes and material waste. The AM process is tool-less, hence design iterations are quickly implemented and validated through changes in virtual CAD files. The Aerospace and MedTech industries are currently at the forefront of industry adaption. Especially in the aerospace industry, high cost-to-fly parts with small build volume can benefit from AM technology. Equally, as medical devices move to bespoke design rather than mass-produced standard sizes, the product demand for AM increases. Lastly, the spare parts market is expected to benefit significantly from AM technologies to reduce cost associated with making, storing and shipping spare parts and hence reduce lead time for international customers.

You also teach at universities. What is your advice for young engineers looking for interesting fields in which to work?

Emerging research and engineering fields open up constantly due to today’s rapid progress in manufacturing technologies, data analytics and process control. In my opinion, it is therefore important to keep an open mind and be willing to change and adapt research interests throughout one’s career to embrace new challenges. The expertise gained from working in different fields and application areas will always be useful for future projects.
You’ve written a great number of papers in a wide variety of engineering disciplines. What are the benefits you’ve experienced from collaborating with others and publishing the results of your work?

I have worked on large international projects and collaborated with research institutes and universities in the UK, Switzerland, Australia and the US. In my opinion, international and crossdisciplinary collaboration is crucial to progress research and develop original contributions. Collaboration also has the advantage of utilizing unique experimental set-ups and equipment at other research institutions to widen the scope of possible research. Publication of low TRL research outcomes is very important to allow for the dissemination and uptake of novel ideas and outcomes in future technology development.

For more information, visit Singapore Institute of Manufacturing Technology.

This article was published in the September 2017 issue of SIMULIA Community News magazine.